In spite of the fact that marital instability is common among people in their mid years, most of the information we have about the causes of divorce and permanent separation is based on the experiences of young persons. As we have shown above, there is good reason to believe the causes of marital instability midway in the life course are quite different than they are for young people. Using a national sample of married persons we propose to identify those differences. A longitudinal design will enable us to directly link changes in such independent variables as economic resources, wife employment, presence of children, marital satisfaction, life goals and health to martial instability. The panel design of the study will enable us to separate age and family life cycle effects and to assess the way in which changes in the proposed casual variables influence the stability of the marriage. We define marital instability as a process that includes the full gamut of activities from thinking about divorce and discussing it, to actually filing for separation or divorce, to obtaining the decree. Part of the significance of the study rests on the fact that information on causes of marital instability for people in different stages of the life course will be of value to family counselors and other health care specialists who deal with people whose marriage may be faltering. Information from the study will also help in making more accurate estimates of the number of people who will divorce or separate which will be of assistance in anticipating needs for a wide variety of health and human services.